Stop sorting résumés. Measure the fit.
Cevio replaces PDF résumés with a structured questionnaire, ranks every application with a score explained criterion by criterion, and leaves the final decision to you.
- Free during the test phase
- Jobs and applications in a single console
- No automatic rejection
The role
Senior Fullstack Engineer
Fit
8
The candidate
Sarah Benali
Weak fit. Time-zone conflict, experience off target. Decision: yours.
- Questions
- 30at most, written and weighted by you
- Score
- 0–100broken down, never a black box
- Auto-rejections
- 0a human always decides
- Résumés to parse
- 0no PDF to reread
Interactive demo
This is what your team will see. Try it.
The candidates below are fictitious; the screens and calculations are the tool's own. Open an application, read the score criterion by criterion, then make the decision yourself.
Senior Fullstack Engineer/Acme Technologies/5 applications
Ranking
by Jev scoreSarah Benali
93.4
/ 100
How this score is calculated
Base 94.9×(0.20 + 0.80 × 0.98)=93.4(×0.98)
Declared constraints
Based in Paris (75011). Eligible to work in France/EU, no visa required. Available with a negotiable 1-month notice. Expectations: €72k gross per year. Open to hybrid work (2 days office / 3 days remote).
Jev ranks and informs. No profile is rejected without your action.
Score transparency
A score you can redo by hand.
No black box: weighted criteria, a probability of eligibility, a public formula. Here is the default grid: move the sliders and check for yourself.
The formula
Base = Σ (poids × note ÷ 4) × 100
Score = Base × (0.20 + 0.80 × Peligible)
A strict conflict (time zone, work permit, salary) removes up to 80% of the score by default, without ever deleting the profile. Weights and discount can be set for each job.
Overall score
67.7/ 100
Your evaluation grid
Your questions, your criteria, your ranking.
A tech scale-up and an international customer service team don't hire on the same priorities. With Cevio, each company writes its questionnaire and weights its criteria: Jev evaluates on your grid, not on a one-size-fits-all grid.
Your own questions
Up to 10 steps and 30 questions: free text, short answer or single choice. Adaptable job by job.
One weight per criterion
Every scored question becomes a criterion, with your evaluation guidelines and, if you wish, your own 0-to-4 grid.
Eligibility or plain information
Mark blocking constraints, or keep an answer to yourself: salary expectations are never passed to the model.
Instant re-ranking
Change a weight: applications already received are re-ranked immediately, with no new evaluation.
Example grids
The default, balanced grid, before any customization.
Ranking · 4 applications
Change vs Cevio grid
- Léa Roussel, Balanced profile, very motivated: 80.3 out of 100
- Inès Moreau, Technical expert, reserved people skills: 77.9 out of 100
- Karim Haddad, Long career, generic motivation: 74.8 out of 100
- Marc Dubois, Bilingual, excellent customer relations: 71.1 out of 100
Same answers, same Jev grades: only the weighting changes. In the console, re-ranking is immediate, with no new evaluation, and candidates never see your grid.
Why Cevio
Neither résumé parsing nor opaque scores: a sort you can defend.
The Jev model (TypeSafe AI) produces calibrated probabilities in a single pass, and every score breaks down. A human always stays in control.
| Criterion | Manual résumé sorting | Keyword ATS | Cevio |
|---|---|---|---|
| Applications comparable with each otherSame questionnaire for every candidate of a job, same fields, same maximum length. | |||
| No fragile PDF parsingInformation is typed in, not extracted from a formatted file. | |||
| Ranking explained criterion by criterionEvery score breaks down criterion by criterion, with a probability of eligibility. | |||
| Criteria and weighting specific to each roleYour questions, your weights: the ranking follows your priorities, not a one-size-fits-all grid. | |||
| Blocking constraints detected earlyTime zone, work permit, salary, availability: flagged, not guessed. | |||
| Final decision human and tracedNo automatic rejection; annotations and model trace kept. |
The candidate experience
Candidates who finish their application.
A short, guided journey with no résumé upload. You receive complete, comparable answers, not heterogeneous PDFs.
- Short steps, designed to be completed in a single session.
- 2,000 characters maximum per field: dense, readable, comparable answers.
- The score is never shown to the candidate: impartial and confidential treatment.
Who are you? Nothing else to provide than a way to reply to you.
Full name
Email address
Phone number (optional)
Aperçu : exemple de candidature fictive.
Features
Everything you need to sort fast, nothing superfluous.
The console is dense and direct: it serves teams that compare dozens of applications.
Structured questionnaire
Your steps, your questions, answers capped at 2,000 characters. Every application for a job arrives in the same format.
Sortable ranking
Sort by overall score, by criterion or by eligibility. Filter by status and search by skill.
Role ↔ candidate view
The role's requirements next to the candidate's answers, in a direct comparison drawer.
Eligibility alert
Strict conflicts apply a visible discount, without ever removing a profile from your list.
Full traceability
Raw model trace, annotations and status of every decision, kept application by application.
Public or private jobs
Publish on the Cevio catalog or share a direct link for a confidential hire.
Custom weighting
Each criterion weighs what you decide. Change a weight: the ranking updates instantly, with no new evaluation.
Isolated workspaces
Every company has its own sealed workspace, protected by a password and a one-time code (2FA).
For candidates
Looking for a job? Apply here.
Pick a job, answer a few short questions, no résumé to prepare. Your application is read on what you write, not on its layout.
No public jobs at the moment
Some companies post privately: use the application link you received from the employer.
Did you receive a direct link from an employer? Open it as is: some jobs are deliberately absent from the catalog.
Pricing
Free during the test phase.
We are opening Cevio to a first group of companies. Access is free for the duration of the test, for a limited time.
Test access
€0during the test phase
Limited time. No credit card required.
- Full recruiter console: ranking, sorting, filters, comparison drawer
- Jev score on your weighted criteria, with eligibility alert
- Custom candidate questionnaire, no résumé
- Public jobs in the catalog or private by direct link
- Annotations, decision statuses and model trace
- Secure sign-in with password and one-time code
After the test
Coming soon
Pricing will be communicated at the end of the test phase, taking into account what we will have learned with the first teams.
Test access is not tied to any payment method.
Frequently asked questions
What we get asked the most.
Do my candidates have to provide a résumé?
No. They answer your job's questionnaire: by default identity, skills, background, motivation, qualities and constraints, and you can add, rewrite or remove questions. You receive structured answers rather than heterogeneous PDFs.
Can Jev reject a candidate on its own?
Never. Jev ranks and flags, you decide. Even in case of an eligibility conflict, the profile stays visible in your list with its details: only a member of your team can shortlist or reject it.
How is the score calculated?
Each criterion gets a grade from 0 to 4; the weights you give them produce a base score out of 100. By default: technical 25%, experience 25%, motivation 20%, writing 15%, attitude 15%. This score is then multiplied by an eligibility factor, between 0.20 and 1 by default. The formula is public, and you can test it higher up on this page.
Can I adapt the questions and criteria to my roles?
Yes. You define a default questionnaire for your company, then adapt it job by job: up to 10 steps and 30 questions (free text, short answer or single choice). Each question can be scored, with its weight, your evaluation guidelines and your own 0-to-4 grid; used to check eligibility; or kept as plain information for you, never passed to the model. Changing a weight instantly re-ranks the applications already received.
What happens if a candidate is brilliant but ineligible?
The score takes a discount of up to 80% (rate adjustable for each job), a warning badge explains why (time zone, work permit, salary expectations, availability), and it is up to you to judge whether the constraint is really blocking.
Do candidates see their score?
No. Scores, weighting and your evaluation guidelines are strictly reserved to the recruiter console, to preserve impartial treatment.
How do I get access, and is it really free?
Fill in the form at the bottom of the page. We send you a secure invitation to create your company workspace (password and one-time code). Access is free during the test phase, which is time-limited; pricing will be communicated afterwards.
I'm a candidate: where can I find the jobs?
In the job catalog, reachable from the menu and from this page. If an employer sent you a direct link, open it as is: some jobs are private and absent from the catalog.
Free test phase
Put Cevio on your next hire.
Tell us who you are: we will send you a secure invitation to create your company workspace.
- Free during the test phase, no credit card
- Create and publish your jobs yourself from the console
- You keep the decision: no automatic rejection
How it works
From application to decision, in three steps.
Every application arrives in the same format. You finally compare profiles that can be compared.
The candidate answers your questionnaire
The steps and questions you chose for the role. No résumé to upload, no fragile parsing, 2,000 characters maximum per answer.
Jev pre-ranks every application
A score from 0 to 100 on your criteria, weighted by your priorities and adjusted by the probability of eligibility. Blocking constraints (time zone, work permit, salary) are flagged, never hidden.
You decide, everything is traced
Put the role and the candidate side by side, annotate, shortlist or reject. Your decision and the model's raw trace are kept for any audit.